Adaptive contour coding method for encoding a contour image in a video signal

Abstract

A contour coding method adaptively encodes a contour image in a video frame based on the size and complexity of the contour image in order to improve the quality of a reconstructed contour image. First, the size and complexity of the contour image inputted are determined. Simultaneously, the contour image is encoded by using a first and a second contour coding methods to thereby provide first and second coded contour data, respectively. And, the first or the second coded contour data is selected as encoded contour image data based on the size and complexity of the contour image.

What is claimed is: 
     
       1. A method for adaptively encoding a contour image of an object included in a video frame of a video signal wherein the video frame is divided into a multiplicity of blocks, each of the blocks having K×L pixels with K and L being positive integers, respectively, and the contour image consists of a plurality of contour pixels, the contour pixels representing pixels located on a contour, said method comprising the steps of: (a1) coding the contour image by using a first and a second contour coding methods to thereby provide first and second coded contour data, respectively, the first coded contour data being capable of representing the contour image more precisely than the second coded contour data;   (b1) detecting the number of contour blocks, each of the contour blocks including one or more contour pixels therein;   (c1) comparing the number of contour blocks with a first predetermined threshold value;   (d1) calculating curvatures of the contour at selected contour pixels to thereby determine a mean curvature based on the calculated curvatures wherein a curvature of the contour at a selected contour pixel denotes an angle change between two lines that pass therethrough, each of the two lines being drawn through said selected contour pixel and another contour pixel located a predetermined number of contour pixels away therefrom in one direction along the contour;   (e1) comparing the mean curvature with a second predetermined threshold value; and   (f1) selecting the first coded contour data if the number of contour blocks is smaller than the first predetermined threshold value and the mean curvature is greater than or equal to the second predetermined threshold value and selecting the second coded contour data if otherwise, to thereby provide the selected coded contour data as encoded contour image data.   
     
     
       2. The method according to claim 1, wherein said first contour coding method is a chain coding method. 
     
     
       3. The method according to claim 2, wherein said second contour coding method includes a polygonal approximation and a discrete sine transform. 
     
     
       4. The method according to claim 3, wherein said mean curvature is defined as: ##EQU2## wherein C M  denotes the mean curvature; M is the number of selected contour pixels; and θ i  represents a curvature at each of the selected contour pixels. 
     
     
       5. A method for adaptively encoding a contour image of an object included in a video frame of a video signal wherein the video frame is divided into a multiplicity of blocks, each of the blocks having K×L pixels with K and L being positive integers, respectively, and the contour image consists of a plurality of contour pixels, the contour pixels representing pixels located on a contour, said method comprising the steps of: (a2) detecting the number of contour blocks, each of the contour blocks including one or more contour pixels therein;   (b2) calculating curvatures of the contour at selected contour pixels to thereby determine a mean curvature based on the calculated curvatures wherein a curvature of the contour at a selected contour pixel denotes an angle change between two lines that pass; therethrough, each of the two lines being drawn through said selected contour pixel and another contour pixel located a predetermined number of contour pixels away therefrom in one direction along the contour;   (c2) comparing the number of contour blocks and the mean curvature with a first and a second predetermined threshold values, respectively;   (d2) coding the contour image by using ea first contour coding method if the number of contour blocks is smaller than the first predetermined threshold value and the mean curvature is greater than or equal to the second predetermined threshold value; and   (e2) coding the contour image by using a second contour coding method if the number of contour blocks is larger than or equal to the first predetermined threshold value or the mean curvature is smaller than the second predetermined threshold value,   wherein the first contour coding method is chosen to be capable of representing the contour image more precisely than the second contour coding method.   
     
     
       6. The method according to claim 5, wherein said first contour coding method is a chain coding method. 
     
     
       7. The method according to claim 6, wherein said second contour coding method includes a polygonal approximation and a discrete sine transform. 
     
     
       8. The method according to claim 7, wherein said mean curvature is defined as: ##EQU3## wherein C M  denotes the mean curvature; M is the number of selected contour pixels; and θ i  represents a curvature at each of the selected contour pixels.